Data visualization applications enable a user to understand a data set visually, including distribution, trends, outliers, and other factors that are important to making business decisions. However, different data sets from different data sources can have unique characteristics (e.g., data distributions), and sometimes characteristics of data sets from the same data source can vary over time. Some data sets have data values over a wide range (e.g., multiple orders of magnitude). When all of the data in the range is positive, an axis of a data visualization can use a logarithmic scale. However, a logarithmic scale does not work when the range of data values includes 0 or negative values.